Fornax OS

Fornax Execution Consulting
GCC

04 / Decide

JUDGE

Execution Verdict System

You've decided something big. JUDGE works out what would have to be true inside your company for that decision to survive contact with reality, tests whether those things are actually true, and shows you where it breaks.

In design
  • UnitOne decision
  • Domains9
  • TestsConditions
  • ReturnsA verdict

Built on the method Fornax Execution Consulting already runs by hand.

01 Reframe

Most failed decisions were not wrong. They were unexecutable.

The board approves it, the numbers work, the market is real, and eighteen months later it has quietly not happened. Nobody can point to the moment it went wrong because there wasn't one. The company simply could not carry it, and that was already true on the day it was signed.

02 Questions

Say the decision is: enter Saudi Arabia.

That may well be correct. JUDGE doesn't have an opinion on it. What it asks is the set of questions that determine whether you can actually do it.

Plate 01 · Decomposition One decision · ten conditions
  1. 01Who owns this expansion, by name?Unevidenced
  2. 02What authority do they actually hold to commit money and hire?Unevidenced
  3. 03Which existing processes have to change, and who signs those off?Unevidenced
  4. 04Which teams become dependent on each other that were not before?Unevidenced
  5. 05Does anyone have spare capacity, or is this on top of a full plate?Contradicted
  6. 06What information will management see, and how quickly?Unevidenced
  7. 07Who finds out first when it starts going wrong?Nobody
  8. 08Who has the standing to stop it?Nobody
  9. 09What happens if the core assumption turns out to be wrong?Unevidenced
  10. 10Which controls must exist before the first dirham is committed?Unevidenced

Illustrative decomposition. Note what the hard ones have in common: none of them are about Saudi Arabia. They're about you.

Conditions five, seven and eight are the ones that end expansions. Not the market, not the pricing. Nobody had capacity, nobody would notice early, and nobody could have stopped it.

03 Mechanism

Five stages, ending in a verdict you can act on.

Plate 02 · The verdict mechanism Cycling
  1. 01 Decision What exactly are you intending to do? Guarantee: your decision is taken as given, never argued with.
  2. 02 Conditions What would have to be true for it to survive? Guarantee: every condition is testable, not a sentiment.
  3. 03 Evidence Which of those are actually true, and how do we know? Guarantee: a claim and evidence are never recorded as the same thing.
  4. 04 Exposure Where does this break, and what happens when it does? Guarantee: the failure path is named before you fund it.
  5. 05 Verdict Can this company carry this decision, as it stands today? Guarantee: not a score. A reasoned answer with what to fix first.

Fix a condition and re-run it. The verdict is not permanent

Stage three is the one that makes it worth doing. Management saying a control exists is a claim. JUDGE records claims and evidence in different columns, which is uncomfortable and is the entire value.

04 Domains

Nine places a decision gets tested.

These are the domains the consulting practice inspects by hand. The formal name is on the left. What it actually asks is on the right, because a heading you have to decode is a heading that failed.

  1. STRUCTStructural Load & IntegrityCan the shape of the company carry this, or does it bend?
  2. AUTHAuthority & Power ArchitectureDoes anyone actually have the power to make this happen?
  3. ACCTAccountability ClosureWhen this slips, whose name is on it?
  4. SOVRSystem SovereigntyDo you control what this depends on, or does someone else?
  5. AMBTAmbition FeasibilityIs this the right size for what you can actually do?
  6. TRUTHDirectional TruthDoes management know what's really happening, or a version of it?
  7. TRANSTransformation AuthorizationHas anyone actually authorised the change this requires?
  8. LOADLoad Allocation & Failure SurvivabilityIf one part of this fails, does the rest hold?
  9. ECONStructural EconomicsDoes the money work at the scale this needs?
05 Boundary

JUDGE never tells you whether your strategy is right.

  • It has no view on whether entering that market is wise. That's your judgement and it stays yours.
  • It won't produce a number that stands in for thinking. A score with no reasoning attached is a way of avoiding the argument.
  • It won't accept that a control exists because someone senior said so. A claim and evidence go in different columns.
  • It isn't a governance layer that slows every decision down. It's for the few that are large enough to hurt.
  • It can't tell you what will happen. It can tell you what has to be true, and which of those things currently isn't.

The distinction is the whole product. Given the strategy you have already chosen, JUDGE tells you whether this organisation, as it exists today, is able to carry it out. That's a narrower question than whether the strategy is good, and a far more answerable one.

06 / Status

The method exists. The software doesn't yet.

Fornax Execution Consulting already runs this analysis by hand, for private equity firms, family offices and capital allocators across the GCC. The nine domains above aren't a product concept, they're the working taxonomy of an existing practice.

JUDGE is the attempt to productise it so an in-house team can run it without hiring the practice. It's in design, there's nothing to buy, and this page will say so until that changes.

If you have a decision in front of you now and can't wait for software, the consulting practice is the answer today.

Ask about JUDGE

Tell us the kind of decision you'd want tested. We'll tell you whether this is the right instrument for it, or whether it's a consulting problem.

Sent

We answer within one business day.

JUDGE asks whether you can start. WATCHTOWER asks whether you're still on course once you have.